Pearson correlation analyses were performed between the different ABAS-II domains, GAC and the results from all cognitive and executive function tests.
To explore any relationships between the results from cognitive / executive function tests and GAC, linear regressions were performed with GAC as
dependent variable and correlating cognitive / executive function measures as independent variables.
disability, was explored for associations to common factors associated with functional level such as age, processing speed and IQ, presence of ADHD, degree of co-occurring psychiatric diagnoses and degree of ASD. Categorical data was analyzed using Chi-squared test or Fisher’s exact test (for 2x2 tables).
Correlation analyses between GAF and potentially associated factors were performed using Spearman’s rho, as all independent variables were not normally distributed. Linear regression analyses were performed for exploring the association between the outcome variable (GAF) and potential predictors.
3.4.3 STUDY III
As in study II, the ASD and Subthreshold ASD groups were collapsed into a
“Merged ASD” group. Degree of ASD symptoms was compared between depressed and non-depressed participants, to assess if ASD diagnoses (unintentionally) were assigned to symptoms of depression or vice versa.
Known risk factors for suicide attempts were analysed as potentially associated to suicide attempts in the study sample. The factors were dependence on economic welfare, age, sex, degree of psychiatric comorbidity, number of depressive symptoms met during most severe depressive episode, experience of psychological trauma, ASD symptomatology, IQ and PSI.
Known risk factors for NSSI were analysed as potentially associated to NSSI in the study sample. Those factors were cluster B personality disorders, ASD symptomatology, sex, age, history of suicidal ideation and/or suicidal behaviour and occurrence of ADHD.
Between-group comparisons were performed with Mann-Whitney U test or Student’s t-test (for continuous variables, depending on if the variables were normally distributed). Chi-squared or Fisher’s exact test (for 2x2 tables) were used for comparisons of categorical data. Relationships between the outcome variables (suicide attempts and NSSI) and the different potential predictors were explored using logistic regression analyses.
3.4.4 STUDY IV
Pearson correlation analyses were performed between the different ABAS-II domains, GAC and the results from all cognitive and executive function tests.
To explore any relationships between the results from cognitive / executive function tests and GAC, linear regressions were performed with GAC as
dependent variable and correlating cognitive / executive function measures as independent variables.
3.5 ETHICS
All procedures performed in the context of these studies comply with the ethical standards stated in the Helsinki Declaration of 1975, as revised in 2008.
Written or oral informed consent was obtained from all participants. The work in this thesis was approved by the Regional Ethical Review Board in Lund, Sweden (Study I, II and III: Reference number 2018/740 and Study IV:
Reference number 2015/696).
4 RESULTS
4.1 STUDY I
Between November 1, 2019, and October 31, 2020, 562 new outpatients (321 females and 241 males; age M = 33.7, Sd = 12.5) were assessed at the psychiatric assessment unit (PAU) in Helsingborg, Sweden. Out of the total 562, 304 (58%; 183 females and 121 males; age M = 33.1, Sd = 12.6) participated in screening for ASD. Out of the screening responders, 197 were screen-positive (65%; 117 females and 80 males; age M = 32.0, Sd = 11.7).
Out of the screen-positive patients, 48 participated (24%; 29 females and 19 males; age M = 32.0, Sd = 11.8) in the study and were subject to in-depth assessment. Out of the participants, 26 met criteria for ASD (54%; 17 females and 9 males; age M = 31.0, Sd = 12.2), 8 (17%; 4 females and 4 males; age M
= 40.6, Sd = 11.8) met criteria for subthreshold ASD and 14 did not meet criteria for ASD or subthreshold ASD (29%; 8 females and 6 males; age M = 28.9, Sd = 9.0).
Assuming (on the basis of similarities in age, sex distribution, reasons for being referred to the clinic and screening results) that the study participants were representative of all screen-positive screening responders and assuming that none of the screening non-responders and none of the screen-negative screening responders had ASD, the prevalence of ASD in this population would be 18.9% - with an additional 5.7% having subthreshold ASD. A maximum estimate, assuming that none of the screen-negative screening responders would have ASD but that the screening non-responders would have a similar distribution of screening results as the screening responders, the prevalence of ASD would be 35% with an additional 10.7% having subthreshold ASD. As an absolute minimum, assuming that we by chance managed to recruit all patients with ASD and subthreshold ASD in the entire population, the prevalence of ASD would be 4.6% with an additional 1.4%
having subthreshold ASD.
4.2 STUDY II
Eighty-one screen-positive patients participated in the study. Patients eligible for inclusion in the control group (RAADS-14 result ≤ 6) proved difficult to recruit into the study, with only nine participating in total. Of the total 90 participants (54 females and 36 males; age M = 31.0, Sd = 10.5), 63 met criteria for ASD (n = 52; 32 females and 20 males; age M = 30.6, Sd = 9.8) or subthreshold ASD (n = 11; 4 females and 7 males; age M = 38.6, Sd = 11.7), together constituting a “Merged ASD” group. Twenty-seven (18 females and 9 males; age M = 29.0, Sd = 10.4) of the participants did not meet criteria for ASD or subthreshold ASD. Based on the distribution of RAADS-14 results (presented in detail in study I), the participants who did not meet criteria for ASD or subthreshold ASD from the original control group (n = 8), together with the participants that did not meet criteria for ASD or subthreshold ASD from the screen-positive group (n = 19; together constituting a “Non-ASD group”), were considered representative of non-ASD patients from the study population.
Psychiatric co-occurring conditions in the Merged ASD group by DSM-5 category, in order of commonness, were: NDDs other than ASD (90%; of which 67% had ADHD of any subtype); Anxiety disorders (87%); OCD and related disorders (53%), depressive disorders (40%); potential SUD (hazardous or harmful alcohol use and/or drug-related problems) (30%); bipolar disorders (29%); trauma and stressor-related disorders (21%); disruptive, impulse- control and conduct disorders (11%); eating disorders (10%); and schizophrenia spectrum and other psychotic disorders (5%). Mean number of comorbid psychiatric diagnoses were 4.1, with a standard deviation of 1.7. No differences were found between the Merged ASD group and the Non-ASD group in self-reported social relationships, living conditions, educational level, vocational status, psychiatric history or history of support in school or from social services. Psychiatric profiles for the two groups were similar, but anxiety disorders and NDDs (other than ASD) were more common in the Merged ASD group. There were non-significant trends in the data towards depressive disorders, obsessive-compulsive and related disorders and trauma- and stressor-related disorders being more common in the Merged ASD group as well. The data suggested that parent-rated developmental history show clear signs of NDD symptomatology regardless of classification (Merged ASD or Non-ASD), though the sample size was too small to be able to compare the groups statistically. The Merged ASD group had more comorbid psychiatric diagnoses and lower GAF scores than the Non-ASD group. Within the Merged ASD group, the number of comorbid psychiatric non-mood disorders and current mood disorder explained 26.7% of the variability in GAF.
4.3 STUDY III
This study explored suicidal thoughts, plans and attempts, as well as NSSI, in the Merged ASD group (n = 63). Active suicidal ideation (actively thinking about ending one’s own life) was experienced by 58% of the participants during the past year. Suicidal plans had been prevalent in 29% of the participants during the past year and 7% of the participants had attempted suicide during the past year. One fourth of the participants had a history of at least one suicide attempt during their life. Some 32% of the participants had engaged in NSSI during the past year, 44% at any time during their life. There was a significant overlap between NSSI and suicidality. The four most common types of NSSI behaviours were hitting oneself on purpose, picking at wounds, cutting or carving oneself and biting oneself. The five most common reasons for engaging in NSSI were relief of feeling numb or empty, to feel something – even if it was pain, to get control of a situation, to punish oneself and to stop bad feelings.
Hazardous or harmful alcohol use and/or drug-related problems and severity of the worst experienced depressive episode explained between one fifth and one third or the variance in suicide attempts. Factors associated with NSSI were being female, ever having experienced suicidal plans and antisocial personality disorder; together explaining up to half of the variance in the occurrence of NSSI. No completed suicides were recorded at the follow-up of medical records. Out of the nineteen participants having a hazardous or harmful alcohol use and/or drug-related problems, only six (four out of six males and two out of thirteen females) had a clinical diagnose of harmful use or addiction at the follow-up of medical records.
4.2 STUDY II
Eighty-one screen-positive patients participated in the study. Patients eligible for inclusion in the control group (RAADS-14 result ≤ 6) proved difficult to recruit into the study, with only nine participating in total. Of the total 90 participants (54 females and 36 males; age M = 31.0, Sd = 10.5), 63 met criteria for ASD (n = 52; 32 females and 20 males; age M = 30.6, Sd = 9.8) or subthreshold ASD (n = 11; 4 females and 7 males; age M = 38.6, Sd = 11.7), together constituting a “Merged ASD” group. Twenty-seven (18 females and 9 males; age M = 29.0, Sd = 10.4) of the participants did not meet criteria for ASD or subthreshold ASD. Based on the distribution of RAADS-14 results (presented in detail in study I), the participants who did not meet criteria for ASD or subthreshold ASD from the original control group (n = 8), together with the participants that did not meet criteria for ASD or subthreshold ASD from the screen-positive group (n = 19; together constituting a “Non-ASD group”), were considered representative of non-ASD patients from the study population.
Psychiatric co-occurring conditions in the Merged ASD group by DSM-5 category, in order of commonness, were: NDDs other than ASD (90%; of which 67% had ADHD of any subtype); Anxiety disorders (87%); OCD and related disorders (53%), depressive disorders (40%); potential SUD (hazardous or harmful alcohol use and/or drug-related problems) (30%); bipolar disorders (29%); trauma and stressor-related disorders (21%); disruptive, impulse- control and conduct disorders (11%); eating disorders (10%); and schizophrenia spectrum and other psychotic disorders (5%). Mean number of comorbid psychiatric diagnoses were 4.1, with a standard deviation of 1.7. No differences were found between the Merged ASD group and the Non-ASD group in self-reported social relationships, living conditions, educational level, vocational status, psychiatric history or history of support in school or from social services. Psychiatric profiles for the two groups were similar, but anxiety disorders and NDDs (other than ASD) were more common in the Merged ASD group. There were non-significant trends in the data towards depressive disorders, obsessive-compulsive and related disorders and trauma- and stressor-related disorders being more common in the Merged ASD group as well. The data suggested that parent-rated developmental history show clear signs of NDD symptomatology regardless of classification (Merged ASD or Non-ASD), though the sample size was too small to be able to compare the groups statistically. The Merged ASD group had more comorbid psychiatric diagnoses and lower GAF scores than the Non-ASD group. Within the Merged ASD group, the number of comorbid psychiatric non-mood disorders and current mood disorder explained 26.7% of the variability in GAF.
4.3 STUDY III
This study explored suicidal thoughts, plans and attempts, as well as NSSI, in the Merged ASD group (n = 63). Active suicidal ideation (actively thinking about ending one’s own life) was experienced by 58% of the participants during the past year. Suicidal plans had been prevalent in 29% of the participants during the past year and 7% of the participants had attempted suicide during the past year. One fourth of the participants had a history of at least one suicide attempt during their life. Some 32% of the participants had engaged in NSSI during the past year, 44% at any time during their life. There was a significant overlap between NSSI and suicidality. The four most common types of NSSI behaviours were hitting oneself on purpose, picking at wounds, cutting or carving oneself and biting oneself. The five most common reasons for engaging in NSSI were relief of feeling numb or empty, to feel something – even if it was pain, to get control of a situation, to punish oneself and to stop bad feelings.
Hazardous or harmful alcohol use and/or drug-related problems and severity of the worst experienced depressive episode explained between one fifth and one third or the variance in suicide attempts. Factors associated with NSSI were being female, ever having experienced suicidal plans and antisocial personality disorder; together explaining up to half of the variance in the occurrence of NSSI. No completed suicides were recorded at the follow-up of medical records. Out of the nineteen participants having a hazardous or harmful alcohol use and/or drug-related problems, only six (four out of six males and two out of thirteen females) had a clinical diagnose of harmful use or addiction at the follow-up of medical records.
4.4 STUDY IV
Results from the cognitive tests showed that the sample assessed (N = 30; 21 males and 9 females; age M = 21.6, Sd = 2.9; eight having additional attention- deficit diagnoses) was of normal but slightly below average IQ. Three of the four measures of executive functions were in the average range, only the Proverb test was below average. Trends in the data showed that PRI from the WAIS was higher than VCI, WMI and PSI. The trends in the data further showed that all tests of executive function except verbal fluency, condition 2 (word-generation in concrete categories), were below average. None of these trends were significantly below the normative means. However, below- average results regarding cognitive flexibility, abstract verbal fluency, inhibition and verbal abstract thinking are often seen clinically in adult psychiatric patients with ASD.
Adaptive function (expressed as GAC) was predicted by IQ, WMI and PSI, respectively. A regression model incorporating WMI and PSI explained 23.8%
of the variance in GAC. IQ alone explained 11.5% of the variance. Mean GAC in the sample was 62.1 (Sd = 15.7), indicating adaptive function in the range of mild ID.
5 DISCUSSION
The studies included in this thesis concerned different aspects of adult psychiatric outpatients with ASD. In study I, we found a relatively high prevalence of ASD and ALTs in newly referred patients to a Swedish psychiatric clinic. In studies II and III, we found that when comparing patients with and without ASD, comorbidity patterns (including suicidal behaviours) were not very dissimilar. In study IV, we found associations between cognitive factors and functional level in patients with ASD.